When Design Becomes Infrastructure

Most organizations treat visual design as window dressing—a final polish applied after the real work is done. The United States Centers for Disease Control and Prevention took a different approach during COVID-19, and the results demonstrate why aesthetics divorced from strategy is merely decoration, while design grounded in purpose becomes infrastructure.

Cesar Rivera, formerly the CDC’s executive creative director, spent the pandemic years answering a question most managers never consider: when millions of lives depend on clear communication, what does “making it look better” actually mean? His answer reveals an uncomfortable truth about organizational competence. Beautiful design that distracts from information is worse than ugly design that clarifies it. During COVID-19, consistency wasn’t about brand vanity—it was about instant recognition in a doctor’s office hallway, where a person needed to identify critical health guidance from fifty feet away.

The challenge wasn’t technical. The CDC already had illustration styles and color palettes. The challenge was human: scientists who wanted comprehensive detail competing with public health officials who needed immediate comprehension, all while serving audiences ranging from healthcare providers to parents to elderly patients with vision impairments. The solution required what Rivera calls “accessible tools”—features designed for people with disabilities that inevitably benefit everyone. Speed-reading audio for English speakers becomes language comprehension support for Spanish speakers. High-contrast text for the vision-impaired becomes clarity for anyone reading in bright sunlight.

This pragmatism extends beyond accessibility compliance. When AI-generated images began appearing in scientific blogs, Rivera’s team rejected them not because they were AI-generated, but because they were wrong. They included things like walls that looked unstable and people with two mouths. These errors would have real consequences, distracting scientists from the information itself. Rivera’s determination wasn’t aesthetic judgment but functional analysis: does this image serve the information or sabotage it?

The broader lesson applies well beyond government agencies. Organizations that treat design as decoration discover too late that their visual communication actively undermines their message. Those that treat design as infrastructure—building systematic approaches to clarity, consistency, and accessibility—create assets that scale across audiences, platforms, and crises. During emergencies, there’s no time to debate brand guidelines. Either your visual system works under pressure, or people misunderstand information when understanding matters most.

The CDC’s approach wasn’t revolutionary. It was simply disciplined: know your audience, eliminate distraction, test against reality, adjust based on evidence. In other words, treat design like any other operational function that determines whether your organization succeeds or fails.

What we’re into this week

Aaron:

I honestly am not sure I could take on the seriousness of communications that Cesar took on a CDC, but I know I could get better at using data in visuals clearly. For this reason, this week I’ve really been into Cocktails, 77 drinks every bar person and party monster should know, a piece that’s a great example of using the right visualization to make navigating through data easy and intelligent.

Scott:

As a designer who’s had a lot of the types of conversations Cesar described, this interview made me think of how designers undercut ourselves way too often. In this video about LED traffic lights, you can see how designers make the mistake of eliminating good solutions, ie, LED traffic lights, because in some edge cases they’re not so good. It’s always got to be a balance between good enough and not good enough; if we only ship perfect, we’ll never ship anything.

Ana:

The different parameters of excellence Cesar navigates in his roles really strike me. The ways that we frame conversations and present material matter as much as the material itself. It made me think of this recent article from AI company OpenAI that tries to explain why generative AIs hallucinate. It turns out: a lot of it has to do with the framing of the test. The AIs and backing teams want to beat the evaluations, but for that motivation to be helpful IRL, the evaluations have to be framed to the real world, not to the lab.

Work across business silos

This week we learned from Cesar Rivera how working across silos of designers, business leaders, scientists, and administrators is incredibly tough.

If you’re looking for a way to get smart on working across creative, business, and specialist teams to recode your corner of America, or do DOGE but do it right, check out Ishmael Interactive’s Team Training. We’ve trained thousands of individual contributors, teams, and C suite leaders in how to truly work across silos and springboard collaboration in their organizations.

We would also love to hear about your experiences working with designers and design teams. Let us know at hello@ishmaelinteractive.com.

Credits

Interviewer: Aaron MeyersGuest: Cesar RiveraProducer: Ana MonroeArticle: John JayArtwork: Sunlight and Shadow: The Newbury Marshes c. 1871/1875 Martin Johnson Heade. Via the National Gallery of Art

Ishmael Interactive is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.